4

3. Research Center for Charged Particle Therapy

Dr. Tsujii received a Ph. D. from Hokkaido University in 1985 for his study onradiation therapy. He has been majoring in radiation oncology since 1969, includingparticle beam therapy at New Mexico University, PSI, Tsukuba University andNIRS. He received Scientific Award from Princess Takamatsu Cancer ResearchFund in 2005 and NISTEP Award from National Institute of Science and TechnologyPolicy in 2006. He has been an honorary membership of ESTRO since 2001 and aCoordinate Member, Science Council of Japan since 2006. He has been a Directorof Research Center for Charged Particle Therapy, NIRS since 2003.

Hirohiko Tsujii, M. D., Ph. D.Director of Research Center forCharged Particle Therapy

The Research Center for Charged Particle Therapy(hereafter, abbreviated as "Center") was establishedin 1993 when the NIRS completed construction of theHIMAC. Since then it has been carrying out clinical,biological and physics research using heavy ionsgenerated from the HIMAC. After accumulating clinicalexperiences of carbon ion radiotherapy in various typesof malignant tumors, the Center was successful inobtaining approval from the Ministry of Health, Welfareand Labor for "Highly Advanced Medical Technology" in2003. Thus carbon ion therapy has meanwhile achievedfor itself a solid place in general practice. The HIMAChas been also served for >500 researchers as amulti-user utilization facility for medical, biological andphysics research.

In 2006, when the second Mid-Term of the NIRS wasinitiated, the Center was reorganized to conduct lifescience research on ionizing radiation, focusing oncarbon ion radiotherapy. This would eventually contributeto the improvement of the quality of life of humanbeings. Research plans for the fiscal year of 2006include : clinical study on carbon ion radiotherapy forlocally advanced tumors ; development and improvementof radiotherapeutic techniques ; design study and R&Dfor a new extension of the treatment rooms for theHIMAC ; research on diagnostic imaging ; QA/QC forradiotherapy and radiation protection ; radiobiologicalexperiments for improvement of radiotherapy ;exploration of variability of radiation sensitivity byinvestigating the SNIPs ; research on HiCEP.

The Center is organized of 6 research groups for twomajor topics (A, B and C). Progress of research foreach topic is summarized.

A. Research on the use of heavy ion beams forcancer radiotherapy.

① Development of advanced cancer radio-therapy with charged particle

This subject has been carried out by the ParticleTherapy Research Group (GL; T. Kamada) consistingof 3 teams : Clinical Trial Research Team, ClinicalDatabase Research Team, and Radiation EffectResearch Team.

From June 1994 to February 2007, a total of 3,178patients were enrolled in nearly 50 different phase I/IIand phase II trials and also in Highly Advanced MedicalTechnology of carbon ion radiotherapy. In the year2006, a total of 549 patients with a variety of malignanttumors were treated with carbon ions, in which nearly75 % of the patients were treated in Highly AdvancedMedical Technology.The hypofractionated radiotherapywith employment of larger doses per fraction andshorter overall treatment time as compared toproton or conventional photon radiotherapy has beeneffectively performed. The average number of fractionsper patient was 13 given in 3 weeks. The new MLChaving fine leaves has been developed since 2005 and in2006 the amount of radiation leakage through thisMLC was measured. It was found that the ratio of theleakage dose to the unshielded dose for 400 MeV/ucarbon beams was measured to be about 1 % at theentrance while the currently used MLC gave about 0.6%. The leakage dose decreased as the depth in waterbecame larger. For effective performance of chargedparticle therapy, computer oriented informationsystem is mandatory. In 2006, the Electronic MedicalRecord (EMR) was implemented and coordinationamong several database systems, including theHospital Information System, Therapy Plan Database,Therapy Schedule Management System, PACS andRadiology Information System for Radiation Therapy,was improved. The TCP model proposed by Webband Nahum has been used for analysis of clinicalresults, which provided useful data for clinical practice.

5

Four new protocol studies were initiated in 2006 :chemoradiotherapy of pancreas cancer ; single fractiontreatment of metastatic liver tumor ; short courseradiotherapy of hilar, nodular type NCSLC ; extendedfield radiotherapy of locally advanced cervix cancer.② Development of a novel irradiation system

for charged particle therapyThis subject has been carried out by the Medical

Physics Research Group (GL ; K. Noda) consistingof 4 teams : Accelerator Development ResearchTeam, Irradiation System Research Team, TherapySystem Research Team, and Compact Heavy-IonTherapy System Research Team.

In the fiscal year 2006, research was focused ondevelopment of 3-D scanning method with a pencilbeam for the new treatment facility that was designedas an extension of HIMAC. In this new facility, 3Dscanning method will be used for treatment of both thefixed target and moving target. For this purpose the fastscanning method and phase-controlled re-scanning withgated irradiation were experimentally evaluated. Thefast scanning method was successfully realized bytaking account for extra dose that was measured whenthe spot moves from one position to the next. It wasalso confirmed through computer simulation that thephase-contro l led re- scanning gave a suf fic ientuniformity in both the lateral and depth-dosedistribution even in the moving target. Furthermore,the design study on a rotating gantry system using 3Dpencil beam scanning method was performed. The final90-degree bending magnet is divided into the two for 60and 30 degrees with the scanning magnets being installedbetween them. Total weight of the gantry system wassuccessfully lowered to about 350 tons, which was aboutthe half the weight of the gantry developed at GSI.③ Standardization and improvement of thera-

peutic and diagnostic techniquesThis research covers a wide range of research and

has been performed by the Diagnosis and TreatmentAdvancement Research Group (GL : T. Kamada)consisting of 4 teams: Image Diagnosis Research Team,Image Processing Research Team, Quality ControlResearch Team, and Radiological Protection ResearchTeam.

Image diagnosis research team studied fundamentalaspects on application of new PET tracers for oncologyimaging. This year, tumor hypoxic imaging using62Cu－ATSM was initiated and bone metastasis imagingusing 18F- FNa was also investigated. Image processingresearch team studied a various type of organ motionusing single/serial 4D CT. Quality control researchteam developed a graphite calorimeter for absolutedosimetry in carbon ion irradiation and demonstrated agood linearity of response as a function of absorbeddose. Radiological protection team studied on the doses

given to the patients in X-ray CT examination, in whichTLDs were used for measurement.

B. Research on radiation effects for improvementof radiation therapy

① RadGenomics research concerning theradiation sensitivity

This subject has been carried out by the RadGenomicsResearch Group (GL; T. Imai) consisting of 3 teams :Genetic Information Team, Molecular Radio-oncologyTeam, and Molecular Biostatistics Team.

Normal tissue reactions of cancer patients varyconsiderably after radiotherapy. A numberof observationshave indicated that certain genetic factors play importantroles in this variability. The aim of the RadGenomicsResearch Group is to explore the genetic characteristicsfor both the patient and its bearing tumor, by which thepotentially most effective radiotherapy can be delivered.This, from the molecular-biological standpoint, wouldopen the way to the development of an individual-orientedradiotherapy. In the fiscal year 2006, four researcheswere primarily conducted. First, we developed a noveloptical detection system for on-chip allele-specificprimer extension to conveniently genotype multipleSNPs simultaneously. Second, microarray analysiswith murine tumor models revealed activated molecularpathway with carbon-ion irradiation responsible genesand pathological evidences of superior effectiveness ofcarbon-ion irradiation. Third, using F2 mice descendedfrom two inbred strains, radiation-induced apoptosissensitive C57BL/6JNrs and radiation-induced apoptosisresistant C3H/HeNrs, we identified a significant locuson chromosome 15 for jejunal crypt cell apoptosis. Thisresult will provide useful tools to identify the newradiosensitive loci. Finally, analyzing RAD18-knockout(RAD18-/-) cells generated from human HCT116 cellssuggested a new function of RAD18 for S phase-specificDNA single-strand break repair.

These results will contribute to identify predictivemarkers for individual radiosensitivity for both malignanttumors and surrounding normal tissues. Furthermore,we have established a collaborating network with fiveuniversity hospitals and our Hospital of the Center toallow for "from bench to bedside" research.

② Biological research concerning theimprovement of radiation therapy

This subject has been carried out by the Heavy-IonRadiobiology Research Group (GL ; R. Okayasu)consisting of 4 teams: Biophysics Team, ExperimentalTherapy Team, Cellular and Molecular Biology Team,and Radiation Modifier Team.

Biophysics Team : RBE studies on DNA doublestrand break (DSB) repair indicated indirect effects ofradiation damage even in the cells irradiated with high-LET radiation. Our method of using remaining number

6

of chromatid breaks after irradiation could be used forprediction of an individual radiosensitivity.Experimentson bystander effects with micro beams were started atTIARA/JAERI, PF/KEK, and Spring-8/JASRI.Experiments concerning the inter-comparison of RBEvalues among domestic and/or international ion-beamradiotherapy facilities were completed this year.

Experimental Therapy Team : The studies on a mousemodel of tumor induction revealed that 15 KeV/ mcarbon irradiation gave a lower induction rate thangamma-irradiation. New studies on tumor heterogeneitywere started using the mixture of two tumor types withvarying levels of radio-sensitivity. Among three mixedtumor groups, one showed a different sensitivity ascompared to the control with single tumor type.

Cellular and Molecular Biology Team : Biologicaldifferences between X-ray and heavy ion particle (C,Fe, Ne) irradiation were identified using severalquantitative assays with therapeutic level radiationdoses. Both DNA microarrays and HiCEP analysisdemonstrated some characteristic molecular featureswith high LET irradiation. The mechanism ofradiosensitization by 17-AAG was identified withX-rays, and radiosensitization was also observed withcarbon ions. RNA interference (RNAi) strategy wasused to increase radiosensitivity of tumor cells.

Radiation Modifier Team : One of Vitamin-E analogsshowed a scavenging rate constant three times largerthan that of natural vitamin E. Tocopherol monoglucoside(TMG) and -tocopheryl-N, N-dimethylglycine( -TDMG) showed a significant in vivo radioprotectioneven in post-irradiation administration. An (-Lipoicacid was found to be a good protector for the brain andits functions. For the study of redox- and oxygen-mapping, T1-weighted MRI was shown to have a greatadvantage in evaluating the pharmacokinetics of newlymodified and/or designed nitroxyl contrast agents.③ Transcriptome Research for RadiobiologyThis subject has been carried out by the Transcriptome

Research Group (GL ; M. Abe) consisting of 3 teams :Stem Cell Research Team, Gene Expression Profillingteam, and Model Organism Research Team.

HiCEP is an ideal method for transcriptome analysis,in which the principle is different from hybridization-based methods. This year an automatic HiCEP reactionmachine (HiCEPer) was developed, which permittedto achieve 96 reactions simultaneously within 3 days.This enables us to perform 10,000 reactions per yearand to analyze many applications such as diagnosis,human molecular epidemiology and so on. In order togenerate an assay system for genome reprogramming,we established fibroblast cell lines in which reportergene was inserted by homologous recombination.Stem-cell specific promoter that was identified by uscontrols the reporter gene. This system allows us to

assess candidate genes, because if the candidatesreprogrammed the fibroblast cells, the reporter genewould be expressed. In addition, we performedfunctional analyses of the four genes using theirknockout mice, which were generated last year. Twolines out of the four strains, abnormal chromosomeintegrity, radiosensitivity, oncogenesis and aging havebeen suggested. One out of the four, defects of circadianrhythm and carcinogenesis have been also suggested.The remaining one showed a male infertility. Detailanalysis of their testis demonstrated that spermatogonialstem cells are defective.

Finally, our proposals for medical use of HiCEP weresubmitted to the ethical committee of our institute.Then, protocol study for esophageal cancer was justauthorized and other study using the blood is also underconsideration.

C. Research Project with Heavy Ions at NIRS-HIMAC

Proposals of 121 were accepted and were carried outin FY2006 at HIMAC. The beam time of 5,457 hourswas supplied to those researches.

The 72 papers, 53 proceedings were published, and245 papers were presented at various meetings. Totalof 528 researchers, including 61 foreign researchers,participated in the project.

7

3.1. Developing advanced clinical therapy with charged particle

Dr. Kamada received a Ph. D. from Hokkaido University in 1996 for his studyon radiotherapy of bile duct cancer. He has had 27 years of experience in clinicalresearch on radiation oncology, including 12 years experience in carbon ionradiotherapy at NIRS. Since 2006, he has been group leader of the Particle TherapyResearch Group for developing advanced clinical therapy with charged particle.

: t_kamada@nirs. go. jpTadashi Kamada, M. D., Ph. D.Head, Clinical Oncology

＊Clinical studies to develop therapeutic techniques fordiseases that are difficult to treat with other therapies(such as pancreatic cancer) and for which chargedparticle radiation therapy does not yet have a role.

＊A study on optimizing irradiation methods by diseaseand by region, using clinical investigations oftherapies in which radiation is combined with drugsand operations

＊Development of a comprehensive database on treatment,clinical course and other factors. Comparison andanalysis of domestic and foreign data on particle beamtherapy.

＊To maximize and disseminate the therapeutic effectof charged particle technology, five hundred patientsare to be treated annually. This is the target numbercombining patients taking part in clinical studies andthose receiving high-technology treatments, inconsideration of the fact that the NIRS is primarily aresearch and development facility.

＊The therapeutic effects of treatments developed bythe Institute are evaluated from the viewpoint ofquality of life (QOL) and therapeutic costs. Patients'opinions are collected to gauge their level of satisfactionwith the therapy.

The Particle Therapy Research Group for developingadvanced clinical therapy with charged particle consistsof clinical trial research team, clinical database researchteam, and radiation effect research team. It doesresearch and development on charged particle therapy.Progress of research in each team is summarized.

1) Clinical trial research teamFrom June 1994 to February 2007, a total of 3178

patients were enrolled into clinical trials using carbonion beams generated by HIMAC. Carbon ion radio-

therapy of these patients was carried out by nearly 50different phase I/II or phase II protocols and highlyadvanced medical technology. The number of thepatients in each tumor site treated with carbon ionbeam are listed in Table 1.

Table 1. The number of the patients in each tumor sitetreated with carbon ion beam.

We treated 549 new patients in 2006. Prostate, lung,head and neck, bone and soft tissue, and liver tumorsare the leading 5 tumor types in the trials. A total of2,867 patients who had a follow-up period of 6 monthsor more were included in this report. The clinical trialrevealed that carbon ion radiotherapy provided definitelocal control and offered a survival advantage withoutunacceptable morbidity in a variety of tumors that werehard to cure with other modalities. Using carbon ionbeams, hypofractionated radiotherapy, with applicationof larger doses per fraction and a reduction of overalltreatment times as compared to conventional photonradiotherapy was possible. Carbon ion radiotherapyhas been approved by the Ministry of Health,Labor and Welfare of Japan as "Highly AdvancedMedical Technology (HAMT) " since November 2003.Nearly 75 % of the patients receiving carbon ion

8

radiotherapy were treated by HAMT in 2006.When irradiating a patient with carbon beam, the

patient should be protected from being exposed on anunwanted dose. A multi-leaf collimator (MLC) andpatient collimators are used to spatially limit the carbonbeam for the sake of delivering high localization of thedose to a target. The MLC can easily form an arbitralaperture shape which conforms to a cross sectionalshape of the target by computer control. However,since each leaf of the present MLC is 6.5 mm thick, itis difficult to make the fine shape which is required forthe cases of cancers which are abutting critical organs,such as head and neck cancers. It the case, the patientcollimator is used, which is manufactured by boring anaperture in a brass block, and it takes a couple of daysand cost. Furthermore, use of the patient collimatorhas enforced radiation therapy technologists to set theheavy collimator just above a patient in positioning.Omitting the patient collimator reduces the expenseand the human burden.

A new MLC has been under development since 2005to be applicable to the case in which the patientcollimator is usually required. The MLC is equippedwith 88 pairs of 2.5 mm thick leaf with 0.15 mm spacing.We would like to notify that the thickness is almost 1/3of the present thickness of 6.5 mm. Each leaf has astep-like structure, instead of a tongue-and-groovestructure. The thin multi-leaf, however, gives riseto a problem that the area occupied by the gaps relativelyincrease with respect to the total area. Since the gap isshielded by a half-length leaf, the beam leakage wouldincrease more than the present MLC. The ratio of theleakage dose to the unshielded dose for 400 MeV/ucarbon beams was experimentally proved to be about 1% at the entrance while the present MLC gives about0.6 %. The leakage dose decreases as the depth inwater becomes deeper.

2) Clinical database research teamAt October 2006, we had implemented the Electronic

Medical Record (EMR) and developed a simple inputmethod for the patient's findings, symptom, tumorresponse, and toxic reactions that should be estimatedby the physician during the clinical interview. Weimproved the coordination among several databasesystems (Hospital Information System, Therapy PlanDatabase, Therapy Schedule Management System,PACSs and Radiology Information System for RadiationTherapy). These systems are connected to each otherand data are transmitted to the destination systems. Wecould gather data directly from the information source.We also developed the IHE (Integrating the HealthcareEnterprise) EUA (Enterprise User Authentication)and PSA (Patient Synchronized Application) functionson the existing systems. These functions made it easy

to operate multiple systems. Two PCs (for example :EMR and PACS viewer) are commonly used for theHospital Information System in one clinical unit. Manyphysicians have to enter a user ID and password to loginto these systems. To solve the troublesome manipulation,we developed the function of the IHE-ITI EUA andPSA. We developed middle-ware for the EUA/PSA toreduce the implementation load among the EMR,PACS-viewer, report-viewer, radiation schedulingsys-tem and radiation information system. The EUA/PSA was based on the HL7 CCOW standard and did notsupport multi PCs. So we enhanced the EUA/PSAmechanism for use with several PCs. We realized thatEUA/PSA were essential in a multi-system environment.Our middle-ware resolved the complexities of theapplication implementation. The established guidelinewas useful to unify the user interfaces of eachapplication. We found that the EUA/PSA function willbe inevitable for visual integration.

Among hospitals and/or medical institutions weimplemented the system to share medical data. Thissystem is based upon the IHE Cross-EnterpriseDocument Sharing (XDS) which uses SOAP, ebXMLRIM and Web Service Description Language (WSDL)and HL7. We prepared the Open Source Softwarelicense for the delivery of software. We are nowdeveloping a document source, document repository,document registry and document consumer that weredefined by the IHE XDS. We are planning to open thissoftware made by our project until autumn 2007. Wethink that it is very important to maintain this softwareand to improve them periodically. We are making effortto establish a maintenance framework for open sourcesoftware.

We continued to promote standardization of thedatabase, and the XML module for radiation therapyand to prepare to communicate clinical radiotherapydata with other hospitals and/or medical facilities. Weanalyzed the physician's workflow in radiation oncologydepartments of the Japanese hospitals. In this workflowanalysis, we classified the flow into 4 categories(initiation of radiation therapy, daily treatment,interruption and resume, termination). This categorieswere suitable for Japanese environment.

The NIRS Hospital Information System renewed atOctober 2006 is shown in Figure. 1.

9

Figure. 1 Current status of Hospital Information Sys-tem in NIRS.

3) Radiation effect research team.The RBE model currently used at HIMAC yields a

clinical dose distribution that depends only on LET as afunction of depth while excluding the other factors suchas dose level, tumour type or fraction schedule. Thisprinciple contributes for reducing those unprovenfactors in the methodology and on the contrary enablesto estimate them from clinical results.

The TCP model proposed by Webb and Nahum hasbeen used for the analysis of the clinical results. Theremarkable point of the model is that it takes thevariation in radiosensitivity among patients intoaccount, and was proven to be effective through theanalysis of NSCLC. This year, the analysis wasextended to the following sites : skull-base chordoma,bone and soft tissue sarcoma and rectum cancers. Localcontrol probability of these tumours under 16 fractionswas analyzed with the model. a term in the LQ model,which denotes the radiosensitivity at smaller doselevel, was derived from the analysis for each site. Itwas revealed that the rectum cancer shows similarradiosensitivity with the NSCLC. These sensitivitiesare also close to that of HSG, cultured cell lineoriginated from human salivary gland tumor, whichprovides a standard response to carbon ions in the HIMACRBE model. On the other hand, the rest sites showedslightly higher radiosensitivity than that of the HSG. Itsuggests room for further decreasing unwanted doseexposure to healthy normal tissues surrounding atumour by optimizing dose distribution depending on itsown radiosensitivity. The other interesting findingis that the bone and soft tissue sarcoma tends to becontrolled by less dose than the NSCLC is while thesarcoma is in general considered to be radioresistantagainst conventional X-rays. Difference in the mechanismof biological effect caused by radiations between carbonand X-rays may play a role for this apparent contradiction.

Next to the NSCLC, clinical trials on liver metastasis

from colorectal cancer have been preceded into a singleirradiation. Based on the success of the retrospectiveTCP estimation in the case of NSCLC, predictiveestimation for the case of the liver metastasis fromcolorectal cancer was tried in order to determineappropriate starting dose. Here, the TCP analysis wasperformed using clinical data of the local control withcarbon ions on colorectal cancer. The single fractiondose that corresponds to expected 96 % of TCP levelwas estimated to be 35.0 GyE in the case. Togetherwith the clinical aspects, single irradiation was initiatedwith the fraction size of 36.0 GyE.

The microdosimetric spectra for high-energy beamsof photons, proton, helium, carbon, neon, siliconand iron ions (LET=0.5-880 keV/mm) were measuredwith a spherical-walled tissue-equivalent proportionalcounter at various depths in a plastic phantom. Survivalcurves for human tumor cells were also obtained underthe same conditions. The survival curves werecompared with those estimated by a microdosimetricmodel based on the spectra and the biological parametersfor each cell line. The estimated terms of the LQ modelwith a fixed value reproduced the experimental resultsfor cell irradiation for ion beams with LETs of less than450 keV/mm, except in the region near the distal peak.

1) Ishikawa H, Tsuji H, Kamada T, Hirasawa N,Yanagi T, Mizoe JE, Akakura K、Suzuki H, ShimazakiJ, Tsujii H : Risk factor of late rectal bleeding aftercarbon ion therapy for prostate cancer,

. 66 : 1084-1091,20062) Ishikawa H, Tsuji H, Kamada T, Yanagi T, Mizoe

JE, Kanai T, Morita S, Wakatsuki M, ShimazakiJ, Tsujii H, Working Group for Genitourinary Tumors:Carbon ion radiation therapy for prostate cancer ;results of a prospective phase II study,

81 (1) : 57-64,20063) Imai R, Kamada T, Tsuji H, Tsujii H, Tsuburai* Y,

Tatezaki* S, Cervical spine osteosarcoma treatedwith carbon ion radiotherapy,7 : 1034-1035, 2006

4) Kanai T, Matsufuji N, Miyamoto T, et al. 2006Examination of GyE System for HIMAC CarbonTherapy Int. J. Radiation Oncology Biol. Phys.64： 650-656.2006

5) Kase Y, Kanai T, Matsumoto Y, et al. 2006Microdosimetric Measurements and Estimation ofHuman Cell Survival for Heavy-Ion Beams. RadiationRes. 166 : 629-38.2006

10

3.2. Research on the Next-generation Irradiation System

Dr. Noda received his B. S. degree from the Department of Nuclear Engineering,Kyushu University in 1979. After completing the M. S. program there in 1981,he worked for development of a PET cyclotron from 1981 to 1989, and he alsostudied accelerator physics from 1985 to 1989 in the Institute for NuclearStudy, University of Tokyo. In 1989, he joined the HIMAC project at NIRS, and he was engaged in construction and development of the HIMAC synchrotron.He received his Ph. D . in 1992 from Kyushu University for the study of energy-losscooling. Currently he is Head of Accelerator Development Section, and he holdsthe additional post of Director of the Medical Physics Research Group.

: noda_k@nirs. go. jpKoji Noda, Ph. D., DirectorMedical Physics Research Group

A design study and R&D work on a new treatmentfacility with HIMAC has just been initiated, in order tofurther development of carbon-ion therapy. Thisfacility, which will be connected with the HIMACsynchrotron, will consist of three treatment rooms :two rooms equipped with horizontal and verticalbeam-delivery systems and another with a rotatinggantry. Both the fixed beam-delivery and rotatinggantry systems employ a 3D beam-scanning methodwith the gated irradiation for a moving target as well aswith the irradiation for a fixed target. The treatment inthe new facility can increase its accuracy considerablyand will bring an adaptive treatment.

The new treatment facility will be connected with theHIMAC accelerator complex and has three treatmentrooms. Two of them are equipped with both horizontaland vertical beam-delivery systems and the other isequipped with a rotating gantry. A schematic view ofthe new facility with HIMAC is shown in Fig. 1. Oneof the greatest challenges in this project is to realizetreatment of a moving target by 3D scanning irradiation.In particle therapy, 3D irradiation with pencil beamscanning, which can realize a high irradiation accuracyeven in the case of an irregularly shaped target, hasbeen developed and already utilized for treatment atPSI (Paul Scherrer Institute) and GSI (Gesellschaftfur Schwerionenforschung mbH). However, pencilbeam scanning is more sensitive to organ motionscompared with the conventional broad-beam irradiation.Although the online motion compensation method andthe rescanning method have been developed to addressthis problem in pencil beam scanning, these methodshave not yet been employed for practical use. In 3Dpencil beam scanning irradiation, the interplay effectbetween the scanning motion and the target motion

brings about hot and/or cold spots in the target volumeeven in the gated irradiation, because the size of thedistal and lateral dose profiles of the pencil beam iscomparable to the residual motion range. Therefore,we decided to employ a combination of the rescanningtechnique and the gated irradiation method to avoidproducing hot/cold spots. In order to realize a relativelylarge number of rescannings within an acceptableirradiation time, we carried out our design study in twosteps : 1) conceptual design of a fast scanning system,and 2) simulation of moving target irradiationwith rescanning and gating. The fast scanning strategywas studied with respect to the scanning method, thescanning magnets and their control. Based on theuniform time structure of beam from the HIMACsynchrotron, we developed a novel optimizationtechnique for fast scanning to cut the irradiation time,in which the exposure during transition of each spot istaken into account. We performed simulation studiesof irradiation of a moving target combined with rescanningand the gated irradiation method. We found that thephase-controlled rescanning (PCR) method gave afeasible solution in which the dynamic beam intensitycontrol technique plays an important role to adequatelycontrol the phase correlation under a relatively smallnumber of rescannings. In the PCR method, it isnecessary to adjust the irradiation time for each depthslice to be within 1-2 seconds of the respiration gatewidth. Consequently, we obtained a feasible solutionfor moving target irradiation by our raster scanningmethod with rescanning and gating functions.

In order to prove our strategy of the fast scanningdescribed above, the beam test of the raster scanningirradiation cooperating with the extended flattop of theHIMAC synchrotron was carried out by using the HIMACspot scanning test port. The irradiation control systemwas slightly modified so as to be capable of the rasterscanning irradiation instead of the spot scanning

11

irradiation. In the present stage, we employed the measured dose response of the pencil beam with an energyof 350 MeV/u, which corresponded to a 220-mm rangein water. The beam size at the entrance and the widthof the Gaussian-shaped mini-peak were 3.5 and 4 mm at1 , respectively. The validity of the beam modeland the optimization calculation had been verifiedexperimentally. In the experiment, the spherical targetof 40 mm diameter was irradiated to generate uniformphysical dose field. The total irradiation time wasdecreased to 20 s due to extended flattop comparedwith the fixed cycle operation of 40 s. A cross monitor,consisting of 128 small ionization chambers, wasemployed to measure the dose distribution in the water.The measured dose distribution was compared with thecalculated one, as shown in Fig. 2. The measured dosedistributions were in good agreement with thecalculation result at different penetration depth.Furthermore, it should be noted that there was nodifference of the field quality i. e. the homogeneity andabsolute dose at the center of the field.

On the other hand, we have carried out a design studyof a rotating gantry system with 3D pencil beamscanning as steps toward the construction of a newtreatment facility at HIMAC. Maximum energy andfield size were set to be 400 MeV/u and 150mm square,respectively. Final 90 degrees bending magnet isdivided into two bending magnets of 60 and 30 degreesto install the scanning magnets between them.Although 30 degrees magnet has relatively largeraperture, total weight of the gantry system wassuppressed to be around 350 tons. Furthermore, thephase-space asymmetry compensation method will beemployed by using a thin scatterer foil in the beam line.This technique makes it possible that the beam sizesand distributions in both planes do not depend on therotating angle of the gantry owing to the symmetriccondition.

Fig. 1. : Schematic view of HIMAC and new treatmentfacility

Fig. 2. : Comparison between measured (open circle)and calculated (line) dose distribution.

1. K. Noda et al., "New accelerator facility for carbon-ioncancer-therapy", J. Radiat. Res., 48, 43-54 (2007).

2. T. Furukawa, T. Inaniwa, S. Sato, T. Tomitani,T. Minohara, K. Noda, T. Kanai, "Design studyof a raster scanning system for moving targetirradiation in heavy-ion radiotherapy" , Med. Phys.34, 1085-1097.

3. S. Sato, T. Furukawa, and K. Noda, "Dynamicintensity control system with RF-knockout slow-extraction in the HIMAC synchrotron", Nucl.Instrum. Methods Phys. Res. A 574, 226-231 (2007).

4. T. Inaniwa, T. Furukawa, T. Tomitani, S. Sato,K. Noda, and T. Kana i, "Optimization forfast-scanning irradiation in particle therapy" , Med.Phys., 34 (8) 3302-3311

5. Y. Iwata, S. Yamada, T. Murakami, T. Fujimoto,T. Fujisawa, H. Ogawa, N. Miyahara, K.Yamamoto, S. Hojo, Y. Sakamoto, "Performanceof a compact injector for heavy-ion medical accelera-tors", Nucl. Instrum. Methods Phys. Res. A 572,1007-1021 (2007).

12

3.3. Standardization and Improvement of Therapeutic and Diagnostic Techniques

Dr. Kamada received a Ph. D. from Hokkaido University in 1996 for his studyon radiotherapy of bile duct cancer. He has had 27 years of experience in clinicalreseach on radiation oncology, including 12 years experience in carbon ionradiotherapy at NIRS. Since 2006, he has been Group Leader of the Diagnosis andTreatment Advancement Research Group for standardization and improvement oftherapeutic and diagnostic techniques.

: t_kamada@nirs. go. jpTadashi Kamada, M. D. Ph. D.Head, Clinical Oncology

＊Development of software to create integrated clinicalimages, determine early therapeutic effects andanalyze prognostic factors using a combination ofmultiple diagnostic imaging techniques.

＊Improvement of treatment plans by using integratedimages obtained from advanced dynamic imagingdevices such as 4-dimensional CT.

＊Research and development on indicators of qualitystandards and methods for quality control andassurance of particle beam and photon beam therapiesand of diagnosis using radiation.

＊Advancement and standardization of therapeutic anddiagnostic methods based on investigation of medicalradiation exposure in Japan.

The Diagnosis and Treatment Advancement ResearchGroup for standardization and improvement oftherapeutic and diagnostic techniques consists of theimage diagnosis research team, image processingresearch team, quality control research team andradiological protection research team, and performsresearch into the advancement and standardization ofradiation therapy and diagnostic methods. Progress ofresearch in each team is summarized.

1) Image diagnosis research teamWe studied fundamentals of application of new PET

tracers for clinical diagnosis. The main targets of ourinterests were imaging of cell/tissue metabolicindicators leading to treatment effects. We started tumorhypoxic imaging using Cu-62-ATSM. We assessed thetracer distribution and pharmacokinetic analysis innormal human volunteer in preparation for laterapplying the tracer to heavy ion radiotherapy patients.Activity of blood decreased relatively rapidly and

reached to low level at about 10 minutes after injection.Liver and Urinary system showed very intense activityin Cu-62-ATSM whole body image. Liver activityreached middle or high level about 3 minutes afterinjection and continued to increase gradually.

We were also planning to perform F-18 FNa PETimaging for precise detection and diagnose of bonemetastasis. F-18-Fluoride has higher bone uptakeand faster blood clearance, resulting in a bettertarget-to-background ratio. F-18-Fluoride PET hasbeen shown to be more accurate than Tc-99m-methylenediphosphonate (MDP) bone scintigraphy for thedetection of both sclerotic and lytic lesions in variousmalignancies and was suggested as an alternative tobone scintigraphy, mainly in patients at high risk formetastatic bone disease but also in patients for whomthe detection of metastatic bone disease and its extentis important in selection of treatment, especially foreligibility decision of carbon ion radiotherapy. A workinggroup, we are one of members, for F-18-Fluoride ionPET study was formed in The Japanese Society ofNuclear Medicine and begun its activity in this year.

A method for HIMAC radiotherapy planning usingC-11-methionine PET and MRI/CT fusion image wasdeveloped in cooperation with head and neck oncologistgroup of our hospital for brain and head and neck cancerpatients. Malignant gliomas are the most commonprimary brain tumors in adults. CT and MRI are thestandard diagnosis methods and treatment planning ofmalignant gliomas. PET enables observation of thebiological pathways of tumors, giving additionalinformation about metabolism, physiology and molecularbiology of tumor tissue. One of the most importanttracers for diagnosis of gliomas is C-11-methionine.The image fusions between PET and planning CT(PET/CT) were performed manually with a treatmentplanning system, Pinnacle. Target delineation inPET/CT fusion was used for practical carbon ion

13

radiotherapy.F-18-FLT PET imaging for carbon ion radiotherapy

patients was started in cooperation with DiagnosticImaging Group of Molecular Imaging Center. F-18-FLTis a thymidine derivative that can image tumor cellproliferation. It is considered a good candidate for amarker of therapeutic response. We started to applythis tracer for assessment of carbon ion radiotherapyeffect for patients with lung cancer. Comparativediscussion of F-18-FLT with C-11-methionine will bemade.

A new PET/CT was introduced in our departmentthis year. We have set the environment for clinicalstudy, and we checked its performance according toNEMA NU-2001.

2) Image processing research teamClinical experience with charged particle beam

treatment at several institutions has demonstratedsuperior dose conformation in comparison to photonbeam therapy.

Organs in the thorax and abdomen may movesignificantly during respiration, complicating treatmentin these locations. Voluntary or forced breath-holdtechniques have been proposed to reduce or eliminatethe effects of breathing during both imaging andradiotherapy treatment, but these approaches prolongtreatment and in many cases, are poorly tolerated bypatients. Respiratory motion during treatment resultsin uncertainties, especially, when irradiating with heavycharged particle beams. During respiration, theradiological pathlength can vary as a function of time,as organs move in and out of a given ray, or as thedensity of voxels along the path change. There is a needto appreciate these temporal variations in the planningprocess. We quantif ied range variations due tointrafractional motions (respiratory and heart beat) andinterfractional changes (tumor shrinkage, chest wallthickness, density changes etc.) using single/serial4DCT lung data.

3) Quality control research teamAs for hadron therapy, the team studied both ab

solute and relative dosimetry. Graphite calorimeter wasdeveloped for absolute dosimetry, since the uncertaintyof dosimetry with ionization chamber was relativelylarge due to the uncertainty of w-value for hadronbeams. The calorimeter showed experimentally goodlinearity of response as a function of absorbed dose withuncertainty less than 1%. Regarding relative dosimetrynew multi-layer ionization chamber (MLIC) wasdeveloped for daily depth-dose measurement. Tissueequivalent materials were adopted for the new MLIC toimprove the effect of nuclear fragmentation. Depth-dosecurve measured by the new MLIC showed good

agreement with the depth dose distribution in water.New calculation method was also developed for evaluationof output factor for small field size. The calculation wasbased on empirical formula taken from HIMACexperiment and showed good agreement withmeasurements within ±1 % down to 20mm squarefield. These achievements were applicable to QC andQA for hadron therapy.

As for photon therapy, new treatment planningsystems, XiO and Pinnacle, were introduced to NIRShospital. Before the systems were used clinically, theteam had carried out comprehensive commissioning,which based on international standards such as IAEA,AAPM and ESTRO guidelines. The team were alsoconducting periodic QA system after the commissioning.NIRS is the Secondary Standard Dosimetry Laboratory(SSDL) in radiation therapy. To establish nation-wideexternal audit system for dosimetry in photon therapy,the team carried out pilot study in which postal glassdosimeters were sent to approximately 100 radiationtherapy centers in Japan. Each center was requestedto irradiate 1 Gy to the postal dosimeter in the standardcondition defined the dosimetry protocol. The doseirradiated to postal glass dosimeter was estimatedusing the calibration coefficient of the dosimeter, whichwas determined at SSDL. The pilot study showed 1.6% standard deviation of dose among 100 centers. Themaximum deviation exceeded 5 % for a center. Theteam tried to improve the deviation by consultationwith a personnel of the center. Finally, every centersatisfied 5 % level criteria. Comparative study was alsodone between the glass dosimeter and TLD which hadbeen used as postal dosimeter. The results showed thatglass dosimeters were appropriate for the postal doseaudit with their promising features. The team alsopreformed similar study to Asian countries, China andKorea. The results of two facilities for each countryshowed good agreements within 2 %.

4) Radiological protection research teamAs dose estimations in FY2006, patients' doses of

X-ray CT examinations were estimated by themeasurements using physical anthropomorphicphantoms of adult and child with Thermoluminescencedosimeters (TLDs) encapsulated in glass for severalX-ray CT apparatuses and their conditions of diagnosesin hospitals. Each organ dose was directly measured bydosimeters put in the organ position inside thephantoms, and effective doses were calculated bymultiplying radiation and tissue weighting factors ofICRP Publication 60. There were differences ofestimated doses among X-ray CT apparatuses andconditions. Patients' doses of developing 256-slice CTscanner examinations were also measured andreported.

14

Since the heavy ion radiation therapies performing inNIRS etc. are state-of-the-art technologies in radiology,any regulatory system of their specific radiation protectionfor occupational exposures has not been established inJapan. For considering the propriety to apply currentregulatory system for the therapies, occupationalexposures of radiation workers in domestic heavy ionradiation therapy institutions were estimated by themeasurements of exposures radiated from activatedmaterials of accelerators and phantoms as alternativesof patients in cooperation with researchers of otherinstitutions. The results show that no specificregulation except current one is needed considering themeasured data.

As internal dose calculations, dose estimations ofpatients on nuclear medicine include a lot of uncertaintiescaused by the parameters and calculation modelsthemselves. Using a voxel phantom and Monte-Carlosimulation method, the uncertainties were evaluatedin some injection cases of radiopharmaceuticals. Thedistributions of doses varied depending on theconditions of the injections and patients.

The surveys of medical radiation usage have longbeen continuously performed in our section of NIRS. InFY2006, X-ray CT examinations were selected as thesurvey subject, and a nationwide survey on X-ray CTdiagnoses was done sending questionnaires to hospitalsand clinics possessing X-ray CT. The kinds of diagnoses,frequencies, exposure positions of patients, settingparameters of apparatuses, patients' data such as age,sex etc. were inquired in the questionnaires, and theanalyses of the data of returned questionnaires are inprogress.

The results of the previous survey on annualexamination data of nuclear medicine in 1997 werereported. The total number of examinations on nuclearmedicine was 1.56 million, and collective effective dosewas estimated as approximately 3.33x104 man Sv basedon ICRP Publ. 53 and 80. The analyses of recent surveydata on nuclear medicine are in progress.

1) S. Mori, K. Nishizawa, M. Ohno, M. Endo.Conversion factor for CT dosimetry to assess patientdose using a 256-slice CT scanner,

888-892, 2006.2) M. Matsumoto, K. Nishizawa, K. Iwai, K.

Akahane, T. Maruyama. Nationwide survey ofnuclear medicine practice and estimation of collectiveeffective dose in Japan,75-82, 2006.

3) S. Mori, M. Endo, S. Kandatsu et al, 'A combination-weighted Feldkamp-based reconstruction algorithm forcone-beam CT', 3953-3965(2006)

4) Y. Kusano, T. Kanai, Y. Kase, N. Matsufuji,M. Komori, N. Kanematsu, A. Ito and H. Uchida,Dose contributions from large-angle scattered partic-les in therapeutic carbon beams,

, 193-198, 20075) Y. Kutsutani-Nakamura, S. Sakata, K. Tabushi,

H. Mizuno, T. Ishii, Mohd Moktar bin Nudin, N.Xuan Cu, L. Dong Han, Rafael Cabrigas Solis, Y.Naiguo, L. Apipunyasopon, Chumpot Kakanaporn,K. Hozumi, T. Teranaka, H. Tsujii, Field surveyof Physical QA/QC for intracavitary brachytherapyof the uterine cervical cancer in 7 countries of EastAsia, , Sup. 3, 173-174, 2006

15

3.4. RadGenomics Project for Radiotherapy

Dr. Imai received a Ph. D. from the University of Tsukuba in 1986. Followinga fellowship from the Japan Society for the Promotion of Science for Japanese JuniorScientists at the Institute of Applied Biochemistry, University of Tsukuba, hejoined the Tsukuba Life Science Center, Institute of Physical and ChemicalResearch (RIKEN). From 1988 to 1989, he worked in the Department of Genetics,Washington University Medical School (St. Louis, Missouri, USA) as a visitingresearch associate. Here Professor Maynard Olson sparked his interest in thehuman genome project. After joining The Cancer Institute, (Japanese Foundationfor Cancer Research) in 1991, Dr. Imai worked on cancer and population genomics.He moved to NIRS in 1994. From 2001 to 2006, he worked as the project

leader of the RadGenomics Project. Since 2006 he has been the director of theRadGenomics Research Group.

: imait@nirs. go. jp

Takashi Imai, Ph. D.Director

Cancer patients vary considerably in normal tissuereactions after radiotherapy. Several observations haveindicated that certain genetic factors play importantroles in this variability. It has been hypothesized thatthe clinical radiosensitivity of normal tissues should beregarded as a so-called complex trait dependent on thecumulative effect of many minor genetic determinants.Thus single nucleotide polymorphisms (SNPs) oncertain genes may somehow associate with the severityof normal tissue reactions after radiotherapy. It isimportant to uncover a molecular basis underlyingradiation sensitivity of normal tissues for furtherinvestigation of the more complex character of cancercells. In this study we have searched for polymorphismsthat are associated with normal tissue radiationsensitivity of various cancer patients. We believe theresults will open a way for achieving individual-orientedradiotherapy with high-therapeutic ratio.

The outcome of this research will allow us to identifyany correlations between an individual DNA sequenceand radiation susceptibility (treatment efficiency andadverse effects). If a correlation is found, the DNAsequence in blood cells will enable the prediction of anindividual's radiation susceptibility. Therefore, it willbe possible to provide information to determinetreatment protocols, such as the irradiation method andthe avoidance of adverse effects, leading to personalizedradiotherapy. The project will also contribute to futureresearch on the molecular mechanisms of radiationsensitivity in humans.

PatientsThe 2,090 patients who were registered between

2001 and 2007 included 703 breast cancer patients, 272cervical cancer patients, 461 prostate cancer patients,

and 278 head and neck cancer patients. Normal tissuereactions until the 3rd month after completion of thetreatment were graded according to the National CancerInstitute-Common Toxicity Criteria (NCI/CTC). Lateeffects on normal tissues were graded according to theRadiation Therapy Oncology Group/ the EuropeanOrganization for Research and Treatment of Cancer(RTOG/EORTC) scoring system and the Late Effectsof Normal Tissues-Subjective, Objective, Managementand Analytic (LENT-SOMA) scoring system. Patientswere divided into two groups (radiosensitive andradioresistant) according to the grades determined bythe above scoring systems.

On-chip optical detection system for allele-specificextension of 3'-LNA modified oligonucleotides

SNPs are useful as genetic association markersfor various human diseases as well as for prediction ofindividual responses to therapeutic treatment such asdrugs and ionizing radiation. For routine molecularbiology research and bedside clinical diagnosis, readilyavailable technologies are required to genotype limitednumbers of SNPs that were selected in previous largescale association studies. To this end, easy and rapidprotocols with inexpensive instruments running atreasonable cost are required for the technology to bewidely adopted.

In the present work, a novel optical detection systemfor on-chip allele-specific primer extension has beendeveloped to conveniently genotype multiple SNPs.Optimization of the procedure was achieved by i) lockednucleic acid (LNA) modification of the 3'-end ofimmobilized oligonucleotide primer, ii) titration ofmagnesium concentration of the reaction mixture iii)utilization of optimum reaction temperature. Efficientprimer extension without an annealing step using

16

double-stranded template DNAs was demonstrated forLNA-modified oligonucleotides immobilized on anS-Bio PrimeSurface plastic base. This propertyprovided simplification of experimental procedures andreduction of reaction time to as short as 10 minutesat a constant temperature of 65°C. Incorporation ofbiotin-dUTP during primer extension, followed bybinding of alkaline phosphatase-conjugated streptavidin,allowed optical detection of the typing results throughprecipitation of colored alkaline phosphatase substrateonto the surface of the plastic base. Oligonucleotideprimer sets were designed to genotype three SNPs inthe genes APEX1, TGFB1 and SOD2, previouslyinvestigated for association with radiation sensitivity.The simultaneous evaluation of these SNPs in 25individuals has produced considerably reliable results.

The experimental system developed in this study isnot oriented towards high throughput analysis. Rather,limited numbers of SNPs are easily analyzed within acouple of hours. Dividing the surface of the plastic baseinto multiple areas can increase the number of individualsanalyzed per chip without affecting experimentalprocesses. In conclusion, all the benefits describedabove make this system applicable to routine molecularbiology research and bedside clinical diagnosis.

Microarray analysis of the transcriptional response tocarbon ion irradiation in murine tumors

The purpose of this study was to identify molecularmechanism induced by carbon ion radiotherapy in orderto provide information on potential targets forprediction of its effectiveness. Murine squamous cellcarcinomas, NR-S1 (resistant to gamma-irradiation),and SCCVII (sensitive), were transplanted in hind legsof C3H/He male mice and established solid tumors(7.5-8.5 mm in diameter) were locally irradiated withcarbon ion beamat 30 Gy. Carbon-12 ions were acceleratedby the Heavy Ion Medical Accelerator in Chiba or HIMACsynchrotron up to 290 MeV/u with a dose rate ofapproximately 3 Gy /min. Tumor growth delay (TGD)time, reduction rate of tumor, and recurrence rate oftumor were investigated as parameters of radiosensitivityof tumors. The mice were sacrificed and immediatelydissected before irradiation and after different timepoints, such as 6, 12, 18h, 1, 3, 5, 7, 10, 15,20 days after irradiation or recurrence for transcriptomeassays and pathological investigation. Expressionanalyses were performed using single-color analysismicroarrays consisting of 55k genes. PrincipalCompornent Analysis (PCA) was used to investigatesimilarity of comprehensive overview of the changes ingene expression between expression profiles of twotumors. Analysis of variance (ANOVA) was appliedto the intensity of each tumor at different time point toevaluate significant differences.

Results : TGD time of NR-S1 and SCCVII was 30 daysand 56 days, reduction rate of NR-S1 and SCCVII was40% and 100%, and recurrence rate of NR-S1 andSCCVII was 75% and 50%, respectively. PCA showedthat all expression profiles of NR-S1 were identified asa group, while those of SCCVII were identified asanother group. Recurred tumors showed differentprofiles from non-irradiation control tumors. We detectedgenes, whose expressions were significantly up-regulatedor down regulated at each time point after carbon-irradiation (p value< 0.0001). At 6 hours after irradiation,fourteen genes, which were related with cell cycleregulation, were differentially expressed in bothtumors. Eleven genes, which were related withinflammation or extracellular matrix, were up-regulatedat 6 hours in both tumors, however, their expressionchanges on time-course were different. Pathologicalspecimen showed duplet cells in both tumors 1 day afterirradiation and continuous infiltration of inflammatorycells in SCCVII.

Conclusions : Tumor growth assays revealed that twomurine tumors, which have different radiosensitivityto gamma irradiation, kept their intrinsic radiosensitivityto carbon-ion irradiation. Transcriptional profiling oftwo tumors identified a number of carbon-ion irradiationresponse genes in murine tumors. We have alsoidentified genes as being candidates for predictivemarkers of radiosensitivity to carbon-ion therapy.

Correlation between single nucleotidepolymorphisms and jejunal crypt cell apoptosisafter whole body irradiation

To identify loci concerned with radiosensitivity in amouse model using SNP markers. We subjected276 second filial generation (F2) mice descendedfrom twoinbred mouse strains, radiation-inducedapoptosissensitive C57BL/6JNrs (B6) and radiation-inducedapoptosis resistant C3H/HeNrs (C3H), to 2.5 Gywhole-body irradiation. We quantified jejunal cryptapoptosis, performed a genome-wide survey, andidentified quantitative trait loci (QTL) associated withradiation sensitivity. We expressed apoptosis levelsas an apoptotic score (AS), which was equal to thenumber of apoptotic bodies divided by the number ofcrypts. We genotyped the mice for 109 SNP markers.AS values were 97.7+/-32.9 in B6 mice and 49.0+/-24.9in C3H mice (p < 0.01). Genome-wide analysisrevealed 8 markers (2 on chromosome 9, 4 on 15, 1on 17, and 1 on 18) affecting radiation-induced jejunalapoptosis with log odds (LOD) scores ranging from2.11+/-3.91. We found a significant locus on chromosome15, which was previously reported by Weil and colleagues.These findings support the view that the radiosensitivityof clinically normal tissue depends on variations in severalgenes.

17

Human RAD18 is involved in S phase-specificsingle-strand break repair without PCNAmonoubiquitination

Switching from a replicative to a translesion polymer-ase is an important step to further continue on replica-tion at the site of DNA lesion. Recently, RAD18 (aubiquitin ligase) was shown to monoubiquitinateproliferating cell nuclear antigen (PCNA) in cooperationwith RAD6 (a ubiquitin-conjugating enzyme) at thereplication-stalled sites, causing the polymeraseswitch. Analyzing RAD18-knockout (RAD18-/-) cellsgenerated from human HCT116 cells, in addition to thepolymerase switch, we found a new function of RAD18for S phase-specific DNA single-strand break repair(SSBR). Unlike the case with polymerase switching,PCNA monoubiquitination was not necessary for theSSBR. When compared with wild-type HCT116 cells,RAD18-/- cells, defective in the repair of X-ray-inducedchromosomal aberrations, were signif icantlyhypersensitive to X-ray-irradiation and also to thetopoisomerase I inhibitor camptothecin (CPT) capableof inducing single-strand breaks but were not sosensitive to the topoisomerase II inhibitor etoposidecapable of inducing double-strand breaks. However,such hypersensitivity to CPT observed with RAD18-/-cells was limited to only the S phase due to the absenceof the RAD18 S phase-specific function. Furthermore,the defective SSBR observed in S phase of RAD18-/-cells was also demonstrated by alkaline comet assay.

1. Y. Michikawa, K. Fujimoto, K. Kinoshita, S .Kawai, K. Sugahara, T. Suga, Y. Ootsuka, K.Fujiwara, M. I w a k a w a , T. I m a i : R e l i a b l ea n d Fa s t Allele-Specific Extension of 3'-LNAModified Oligonucleotides Covalently Immobilizedon a Plastic Base, Combined with Biotin-dUTPMediated Optical Detection, 1537-1545,2006

2. F. Wang, Y. Saito, T. Shiomi, S. Yamada, T. Ono,H. Ikehata : Mutation spectrum in UVB-exposedskin epidermis of a mildly-affected Xpg-deficientmouse, , 107-116, 2006

3. S. Nakajima, M. Mori, T. Shiomi, A. Yasui, et. al:Replication-dependent and -independent responsesof RAD18 to DNA damage in human cells,

281, 34687-34695, 20064. N. Shiomi, M. Mori, H. Tsuji, T. Imai, H. Inoue,

S. Tateishi, M. Yamaizumi, T. Shiomi : HumanRAD18 is involved in S phase-specific single-strandbreak repair without PCNA monoubiquitination,

e9, 20075. M. Iwata, M. Iwakawa, S. Noda, T. Ohta, Y.

Minfu, T. Kimura, H. Shibuya, T. Imai: Correlation

between single-nucleotide polymorphisms andjejunal crypt cell apoptosis after whole bodyirradiation, : 181-186, 2007

18

3.5. Biological Research Concerning the Improvement of Radiation Therapy

Dr. Okayasu received his Ph. D. in radiation biology from Colorado StateUniversity, USA in 1987 and worked as a post-doctoral fellow at Thomas JeffersonUniversity, Philadelphia and MD Anderson Cancer Center, Houston. Then hetook a position at Columbia University as an associate research scientist and movedto the University of Texas Medical Branch at Galveston in 1995 as an AssistantProfessor and then onto Colorado State University. In 2002, he moved back toJapan to become a team leader at International Space Radiation Laboratory (ISRL),NIRS and in 2005 he was appointed as Director of ISRL. In 2006, he changed hissection to Research Center for Charged Particle Therapy and became Director ofHeavy-Ion Radiobiology Research Group.

: rokayasu@nirs. go. jp

Ryuichi Okayasu, Ph. D.Director

There are three mid-term plans for this group. Theseare 1) To provide biological experimental data foranalyzing clinical data with regard to tumor control ratioand normal tissue responses for various radiationtherapy protocols, 2) To estimate the risk and benefitratio between tumor cell killing and normal tissuesparing by theoretical calculations based on patients'dose distribution as well as experimental data on celland animal studies. To propose more efficient radiationtherapy regimen by comparing heavy ion radiotherapyand other radio-therapy protocols such as use of X-rays,and 3a) To explore radio-sensitizers and protectorswhich could be used with heavy ion radiotherapy, 3b)To elucidate the mechanism of effective heavy iontreatment for hypoxic tumor cells which show strongresistance to radiation, 3c) To study the indirect(bystander) effects of radiation which occur in non-irradiated cells adjacent to irradiated cells, 3d) Tointegrate the above proposals to improve radiationtherapy and accumulate the biological data resourcesfor a new cancer therapy. These objectives are studiedby four teams including 1) Biophysics Team, 2)Experimental Therapy Team, 3) Cellular and MolecularBiology Team and 4) Radiation Modifier Team. Eachteam has different objectives, however, co-operationsamong four teams are sought in order to accomplish thegoals of this group.

Concerning the modification of RBE values, we foundthat the repair efficiencies with and without existenceof oxygen are different in DNA double strand breaks(DSB) induced by X-rays or carbon ions. These dataindicate that there are indirect effects of radiationdamage even in cells irradiated with high-LET

radiation. The RBE spectra for chromosome aberrationsand remaining DNA damage after repair process are notonly LET dependent but also dependent on the kinds ofaccelerated ions. The RBE of cell killing at very high-doseregion that could not be estimated by calculation wasobtained experimentally. We investigated theradiosensitivity and the remaining number of chromatinbreaks using cells derived from a cervical cancerpatient. These studies eventually lead to the undertakingof carbon therapy for a patient with osteosarcoma andgenomic instability. This is a good example of applicationof radiation biology for tumor treatment. Experimentson bystander effects with micro beams were started atTIARA/JAERI, PF/KEK, and Spring-8/JASRI.We found bystander effects in normal human cellsirradiated with low energy carbon ions. These effectsseem to depend on the number of particles ofaccelerated ions . Exper iments concerning theinter-comparison of RBE among domestic and/orinternational ion-beam radiotherapy facilities werecompleted.

The mouse legs were locally irradiated once withcarbon ions with various LET values, and the effect ofLET on tumor incidence was studied. Life timeobservations indicated that there was no difference inthe induction of tumors between gamma-ray and carbonirradiated mice. Our data also demonstarted the highertumor inductions with the higher radiation doses.Furthermore, to our surprise, the tumor inductionwith 15 KeV/ m carbon irradiation is lower than thatwith gamma-irradiation.

New studies on tumor heterogeneity were startedusing the mixture of two tumor types with varyingradio-sensitivity. These mixed tumors were implantedon the leg of mice and the tumor sensitivities were

19

compared with those of mice implanted with one tumortype. Among three mixed tumor groups, one showeda different sensitivity when compared to the controlwith single tumor type. These studies indicate the im-portance of tumor heterogeneity and warrant furtherstudies.

Biological differences between X-ray and heavy ionparticle (C, Fe, Ne) irradiation were investigatedusing some quantitative assays (H2AX and PCC),focusing on the molecular mechanism in the earlyresponses of DNA damage at the therapeutic levelradiation doses. Comprehensive gene expressiontechniques were employed to several human cell lineswhich were irradiated with X-rays and carbon ionparticles at the doses of therapeutic relevance. BothDNA microarrays and HiCEP, a novel gene expressionprofiling technique developed in our institute, successfullydemonstrated some characteristic feature of themolecular signatures to those different types of ionizingradiation (IR).

One of the potent radio-sensitizers, 17-AAG wasintensively studied in our laboratory. We demonstratedthat 17-AAG could enhance the radio-sensitivity insome cancer cell lines irradiated with X-rays as well ascarbon ions. In this fiscal year, we obtained significantevidence that 17-AAG inhibits the repair of DNA DSBs,especially homologous recombination repair, inducedby IR. Yeast extract is a potential radio-protector whoseeffect was ever shown in vivo (mice). In order to findout the cellular and molecular mechanism for thisprotective effect, in vitro study has just started in ourlab using cell lines treated with the yeast extract. Wealso studied the function of specific DNA DSB repairproteins using RNA interference strategy. For example,BRCA siRNA was found to enhance the radiationsensitivity in HeLa cells by inhibiting repair of DSBs.

The radiation modifier team has studied threesubjects and obtained following results.1) In order to develop better compounds for free radicalscavenging, several novel compounds of vitamin E analog,containing (-chromane ring structure as well as basicpyridine moiety, were synthesized. By the kineticstudy of their free radical scavenging reaction,we found that one of these compounds has a scavengingrate constant three times larger than that of naturalvitamin E.2) For the study of radioprotector, two vitamin Eanalogs were examined. Both tocopherol monoglucoside(TMG) and -tocopheryl-N, N-dimethylglycine ( -TDMG) showed significant in vivo radioprotectionactivity against the lethal dose of whole body X-irradiation

(7.0-7.5 Gy). It is interesting that both compoundsshowed the radiation protection effect even bypost-irradiation administration. Another compound,

-Lipoic acid, was found to be a good protector for brain.The cognitive dysfunction of mice caused by X-irradiationwas ameliorated by the administration of -lipoic acidbefore irradiation.3) For the study of redox- and oxygen-mapping, theresolution and signal to noise ratio of EPR imaging andT1-weighted MRI were compared using an identicalphantom. Several solutions of nitroxyl contrast agentswith different EPR spectral shapes were tested.T1-weighted MRI can detect nitroxyl contrast agentswith a complicated EPR spectrum easier and quicker ;however, T1-weighted MRI has less quantitative abilityespecially for lipophilic nitroxyl contrast agents,because T1-relaxivity, i. e. accessibility to water, isaffected by the hydrophilic/hydrophobic micro-environment of a nitroxyl contrast agent. The less quantitativeability of T1-weighted MRI may not be a disadvantageof redox imaging, which obtains reduction rate of anitroxyl contrast. Therefore, T1-weighted MRI has agreat advantage to examine the pharmacokinetics ofnewly modified and/or designed nitroxyl contrastagents.

1) K. Ando, S. Koike, A. Uzawa, N. Takai, T.Fukawa, Y. Furusawa, M. Aoki, R. Hirayama :Repair of Skin Damage During Fractionated Irradiationwith Gamma Rays and Low-LET Carbon Ions.

167-174 (2006).2) M. Suzuki, C. Tsuruoka, T. Nakano, T. Ohno,

Y. Furusawa, R. Okayasu, The PCC assay can beused to predict radiosensitivity in biopsy culturesirradiated with different types of radiation.

1293-1299 (2006).3) M. Noguchi, D. Yu, R. Hirayama, Y. Ninomiya,

E. Sekine, N. Kubota, K Ando, R. Okayasu :Inhibition of Homologous Recombination Repair inIrradiated Tumor Cells Pretreated with Hsp90 Inhibitor

17-Allylamino-17-demethoxy geldanamycin.351, 658-663 (2006).

4) K. Anzai, M. Ueno, A Yoshida, M. Furuse, W.Aung, I Nakanishi, T. Moritake, K. Takeshita, N.Ikota, "Comparison of Stable Nitroxide, 3-Substituted2,2,5,5-Tetramethylpyrrolidine-N-oxyls, withRespect to Protection from Radiation, Prevention ofDNA Damage, and Distribution in Mice",

1170-1178 (2006).5) K. Manda, M. Ueno, T. Moritake, K. Anzai,

"Radiation-induced cognitive dysfunction and cerebellaroxidative stress in mice : Protective effect of ? -lipoicacid", 7-14 (2007).

20

3.6 Transcriptome Research for Radiobiology

Masumi Abe, Ph. D.Director

Long-range objectives of the 2nd 5-year project(2007-2011)1) Establishing a high throughput system of HiCEP

analysis for dealing with a large number of samplessuch as clinical samples.

2) Developing a HiCEP protocol for blood analysis.3) Developing an assay system for genome reprogramming.4) Functional study using gene knockout mice, in which

genes identified by transcriptome analysis weredisrupted.

For 1), we developed an automatic HiCEP reactionmachine (designated as HiCEPer), which achievessimultaneous 96 reactions with 3 days. This enables usto perform 10,000 reactions per year and to apply theanalysis for many applications such as diagnosis, humanmolecular epidemiology and so on. In addition, weimproved the system of capillary electrophoresis, sothat 48 simultaneous runs for HiCEP analysis becamepossible.

For2), our proposals for medical use of HiCEP havebeen discussed by the ethical committee of ourinstitute. Then, that for esophageal cancer was justauthorized and the other for blood use is underconsideration.

For3), genome reprogramming occurs some particu-lar situations such as when the oocyte nucleus was re-placed by the nucleus of differentiated cell, or whendifferentiated cells were fused with stem cells such asES cells. In both cases, the genome status underlyingthe differentiated cell converts to that underlyingundifferentiated cells, which have a pluripotency.

In order to establish an assay system for genomereprogramming, we attempted to prepare a differentiatedcell line in which reporter gene is existed justdownstream of stem-cell specific promoter. The

reporter will express only when their genome wasreprogrammed by certain stimulus such as genetransfection. More details, we generated lacZ-knockinES cells by means of the homologous recombinationtechnique, in which reportergeneiscontrolledbystem-cellspecific promoter that was identified by us. Withthis ES cells, we generated a knockin mice, fromwhich we generated fibroblast cell lines. Thus, thissystem allows us to assess candidate genes, becauseif the fibroblast cells were reprogrammed by thecandidates, the reporter gene would be expressed.

In addition, we performed functional analyses of thefour genes utilizing their knockout mice, which were generated last year. Two lines out of the four strains,abnormal chromosome integrity, radiosensitivity,oncogenesis and rapid aging have been suggested. Oneout of the four, defects of circadian rhythm andcarcinogenesis have been also suggested. The remainingone showed a male infertility. Detail analysis of theirtestis suggested a severe defect in spermatogensis.

We have developed a new gene expression profilingmethod called HiCEP (High Coverage gene ExpressionProfiling), whose principle is different from that of DNAmicro-array technology. So far fundamentals of thetechnology have been achieved; now, we are currentlyfocusing on the development of a high throughputsystem for the analysis and of the protocol for the analysisusing a small amount of starting materials. This projecthas been supported by Japan Science and Technology(JST) for 5 years.

21

1) Developing an automatic HiCEP reaction machine(designated as HiCEPer)2) Developing a procedure for the analysis using a smallamount of starting materials : 0.5-1.0 ng of total RNAwhich is corresponded to 50 to 100 eukaryotic cells.

1) We have been developed an automatic HiCEPreaction machine. This year we finally achieved theexpected performance. 96 reactions can be performedsimultaneously with 3 days by HiCEPer, enabling us toconduct 10,000 reactions per year. A durability testshowed some problems but all of them have beenovercome.2) Developing a procedure for the analysis using a smallnumber of cells :We succeeded in HiCEP analyses using 0.5-1 ng of totalRNA. During this study, we faced a contamination ofother unknown organisms into the reaction mixture.Especially under 100 cells analysis, this problembecame severe. Even without any RNA, we detectedthe peak pattern clearly. Cloning of the peaks andfollowing sequencing of them disclosed that thesepeaks come from microorganism like psuedemonus. Wechecked all reagents in the HiCEP reaction andreplaced some reagents in which contamination wassuggested. Now, we overcome the contaminationproblem ; therefore it became possible to examine thereaction condition for the HiCEP analysis using lessthan 100 cells. 100-cells HiCEP analysis allows us toconduct transcriptome analysis following FACS sortingor LMD (Laser Microdissection).